Re: [PATCH][RFC] per isr in_progress markers

[Alan Cox]

On Sun, 2002-09-08 at 11:57, Zwane Mwaikambo wrote:
> iirc IDE is capable of doing its own masking per device(nIEN) and in fact 
> does even do unconditional sti's in its isr paths. So i would think it 
> would be one of the not so painful device drivers to take care of. 

If I remember rightly nIEN doesnt work everywhere. Also many IDE
interfaces may be using legacy IRQ wiring rather than PCI irq lines.